Synapse pathology is common to a variety of neurodegenerative and neurodevelopmental diseases. For many such disorders, including schizophrenia and autism spectrum disorders, changes in synapse composition are suspected to be causally related to some of the most devastating symptoms. Synapses throughout the brain share key structural proteins evidenced by their similar, readily identified structure as see through an electron microscope, but they are also very heterogeneous: only a handful of synapse components is known to be shared by all synapses. Presynaptic membranes adhere to postsynaptic membranes in an exceptionally strong interaction, resistant to disassembly. This interaction can withstand tissue fractionation and has been exploited over the past several decades to purify synaptosomes: pre- to postsynaptic adhesions and attached membranes that re-seal to make an enclosed synaptic organelle. Current proteomics studies have utilized synaptosomes or subfractions of such preparations to assess synapse composition. While this supports the feasibility of using synaptosomes for proteomics-based comparisons, the resulting large datasets have not been very useful since they reflect a highly complex starting material containing an enormously heterogeneous population of synapses. Synapse heterogeneity is based largely on differences in connectivity, but currently there are no methodologies available that can be used to interrogate and analyze differences between identified synapse populations. Here we propose a novel and straightforward method based on the mammalian GFP reconstitution across synaptic partners technique that will enable proteomic comparisons between identified synapse populations in health and disease. We have assembled a team to test this using well- defined synaptic circuits. This approach will be a major step forward as it wll permit a rapid, large-scale assessment of particular synapse populations that can be isolated from complex circuits.